Skip to main content

Table 3 Means of the metric values and of their ranks among the different methods over the 14 studied datasets separated into method for the following metrics: diffexpr, skewdiv and corbeaf

From: Combining location-and-scale batch effect adjustment with data cleaning by latent factor adjustment

diffexpr
Mean values combat stand ratioa meanc ratiog none sva fabatch
  0.11044 0.10958 0.10891 0.1088 0.10796 0.10526 0.09517 0.09364
Mean ranks combat stand ratioa meanc none ratiog fabatch sva
  3.28571 3.57143 3.78571 4.14286 4.5 4.64286 5.85714 6.21429
skewdiv
Mean values fabatch sva stand combat ratioa ratiog meanc none
  0.01724 0.02206 0.02377 0.02688 0.02875 0.03257 0.03671 0.05041
Mean ranks sva fabatch combat stand meanc ratioa ratiog none
  2.21429 2.92857 4.28571 4.78571 5.07143 5.42857 5.5 5.78571
corbeaf
Mean values none combat meanc ratioa ratiog stand sva fabatch
  1 0.86857 0.86742 0.85516 0.84931 0.82754 0.69313 0.67795
Mean ranks none combat meanc ratiog ratioa stand sva fabatch
  1 2.92857 2.92857 4.21429 4.35714 5.85714 7.14286 7.57143
  1. In each row the results are listed in descending order according to mean performance in terms of the original values and their ranks, respectively. The results of FAbatch are printed in bold